History of study of quarry pit Q02-2

During a preliminary visit in 2002, I climbed to the Maymeja area using the description and GPS-derived UTM coordinates provided in Burger et al. (1998:205) where I spoke with a herder named Mattias Sarallasi who was working for the land owner from Callalli, Eliseo Vilcahuaman Pambra, and living in the estancia at the base of Maymeja. Sarallasi grabbed a shovel and led me up across the moraines past several small and shallow (1-2m diameter) quarry holes that he indicated to me, and finally to the Q02-2 quarry pit.

In Burger, et al. (1998:204) the authors write that in their 1995 visit Burger and Salas "encountered a local farmer who had been collecting quantities of obsidian from these deposits for future sale." Further, they note that "according to a local farmer, a German resident of Bolivia had visited this obsidian deposit in order to collect obsidian for export to La Paz where it was transformed into craft products". An initial concern for the Upper Colca Project was establishing that the extant pit was not the result of this modern quarrying in the Maymeja area. Directly datable materials were not located in the 2003 research to confirm the antiquity of the pit, but the integrity of the stratigraphy in the test unit, and the presence of cores and bifaces in the excavation, strongly suggest that the quarry pit feature is not a recent excavation. The conclusion here, based on the soil characteristics and the small quarry pits along the ridges below Q02-2, is that recent quarrying work was responsible for the small pile of light soil inside the q02-2 quarry, and for the small quarry pits downslope. The small pile encountered inside the quarry pit feature resembles the level 2 soil encountered in the test unit, suggesting that this modern quarrying entered at least as far as the stratum that was found 15cm below the surface in the 1x1m test unit.

The regional context of this quarry pit, including the relationship between the quarry pit, a Precolumbian road leading to the quarry, and other archaeological features, are described in Section 6.4.1. A discussion of the quarrying methods that may have been used at this location can be found in Ch..

Testing in the debris pile in 2003

In the course of the 2003 Upper Colca Project a 1x1m test unit was placed at Q02-2 in the debris pile immediately downslope of the quarry pit. The unit was aligned to true north and placed on the uphill side of the summit of the debris pile with the aim of bisecting the widest range sizes of pre-existing debris piles as the pile had accumulated below the quarry during ancient use. In the course of the test excavation, the team avoided filling the quarry pit feature with soil or other materials and instead positioned the backdirt piles and screened dirt on the south-east end of the quarry feature.

Figure 7-7. Diagram of Q02-2 quarry pit and position of the Q02-2u2 test unit relative to quarry debris layers and natural slope. By virtue of the quarrying history, levels are inverted in the 1x1m test unit so upper levels came from lower (earlier) levels in the quarry pit.

The quarry pit debris pile included relatively few culturally modified artifacts. While thousands of obsidian "flakes" with conchoidal fracture were encountered, closer inspection revealed that a very small minority were fractured by the human hand. Cortex is sometimes very thin on many of the flakes and nodules at the Chivay source (Section 4.5.1), which complicated the task of differentiating cultural and non-cultural flakes. Non-cultural flakes had no clear bulb or striking platform, or the geometry of the bulb was such that the force responsible for fracturing the rock was evidently far away and was geological in scale. In other words, one would see bulbs of percussion that appeared to have been initiated on adjacent rocks by geological or environmental processes, and the force of the fracture transferred through the obsidian matrix.

Figure 7-8. Q02-2, view of quarry pit and u2 test unit from the north.

Characteristics by level

While the test unit did not expose diagnostic artifacts or datable material, such as carbon, that would have facilitated identifying the culture group associated with the quarrying, it revealed changes in stratigraphy and material that contribute to understanding the prehistory of the Chivay source. Provisional estimates, based on comparisons with dates from the adjacent workshop test unit Q02-2u3, suggest that the bulk of quarrying at Q02-2u2 occurred in the Early Formative. It should be stressed here that because the test excavation is into a debris pile in Q02-2u2, levels are invertedso higher levels in the test unit came from lower in the Q02-2 quarry pit.

Figure 7-9. Q02-2u2 quarry pit south and west profile diagrams showing excavation levels.

Figure 7-10. South profile of Q02-2u2 at quarry pit at the top of levels 11/12.

The test unit Q02-2u2 stratigraphy indicates that episodic filling events occurred in this location. In the profile figure, excavation levels are represented approximately as indicated by strata lines. The profile photo (Figure 7-10) shows the south profile wall, therefore the quarry pit is located off of the left side of the photo frame. As expected, the stratigraphy generally slopes downwards towards the right. However, in level 2 and to a lesser extent in level 3 the stratigraphy levels out and even rises slightly, indicating there was major deposition on the quarry debris pile. These profiles suggest that the quarrying took place in two major episodes with what appears to have been some minor quarrying activity before and after each larger episode. The soil characteristics and artifact evidence provides more insights in to the circumstances surrounding the quarrying.

Table 7-7. Q02-2u2 Excavation levels from test unit at quarry pit. Excavation photos available online.

Table 7-8. Q02-2u2 lithic technical class by test unit excavation level.

The vast majority of obsidian fragments at Q02-2 were non-cultural in origin. The artifact data show that culturally derived flakes were found as low as level 11, and cores and retouched items were found in level 10. These data suggest that the quarry pit area was being exploited for obsidian production with in-situ production before it was being excavated on the scale of the later pit. The distributed nature of the cores, flakes, and retouched artifacts in a stratigraphic profile that is predominantly made up of excavation debris, points to a general focus on procuring whole nodules from this location.

Table 7-9. Q02-2u2 quarry pit obsidian artifact material type and color by level.

A principal question concerning the ancient excavation of the quarry pit is: why excavate for obsidian when it is available on the surface? What desirable characteristics of subsurface obsidian justify the effort expended in digging this quarry pit? Here, the consistency and visual qualities of the obsidian artifacts excavated from Q02-2u2 are examined in order to determine if there were patterns in those characteristics of the material. As shown in Table 7-9, Ob2 obsidian accounted for 7.2% of the artifacts analyzed from the test unit at Q02-2u2, which is comparable to the ratio of projectile point use in the project overall where just 7.8% of the obsidian projectile points observed in the course of 2003 fieldwork were produced on Ob2 obsidian.

Table 7-10. Q02-2u2 Color in quarry pit artifacts compared with all obsidian projectile points from fieldwork.

These data indicate that obsidian from the quarry pit had variability in color of material that was comparable to obsidian use over the entire region. In other words, quarrying individuals at the Q02-2 quarry in the past could expect a relatively low percentage of Ob2 material and color distributions that are consistent with the percentages used for tool production. However, as obsidian was acquired from a number of surface sources, the material quality and color of obsidian from the Q02-2 quarry pit does not appear to have been a principal motivation for excavating in that location. It appears that the quarry pit was relatively typical of the tool characteristics in the region.

However, a difference in cortical flake size from the quarry pit was noted.

Table 7-11. Q02-2u2 quarry pit: Lengths of flakes and cores with >20% cortex.

Because production appears to have been largely conducted elsewhere, the counts for cortical flakes are relatively low. The pattern here suggests that larger nodules were available in deeper in the quarry. Again, levels are invertedin the test unit because the excavation is into a secondary debris pile (Figure 7-7), so higher levels in the 1x1m test unit came from lower levels in the original Q02-2 quarry. It appears that there was an abundance of large nodules and limited flaking was occurring later in the use of the quarry, and therefore larger flakes were being discarded. In a later discussion below, these data are compared with data presented in Table 7-26, from test unit Q02-2u3 at the A03-126 quarry workshop, in order to explore the connection between the quarry pit and the workshop.

A natural slope has been estimated from the surface map created using a total station in 2004 (Figure 7-7). Subtracting the surface map elevation from the calculated natural slope elevation at the location of the Q02-2u2 test unit allows for an estimate of depth at which one would encounter the sterile natural slope in the excavation unit.

Table 7-12. Calculated depth of strata in Q02-2u2.

This independent calculation is confirmed by the Q02-2u2 excavation because sterile soil was encountered at approximately 85 cm depth. Based on the thin strata in the test unit profile, it appears that quarrying in this location was initially low-intensity, but that two episodes of large scale quarrying appear to have occurred in level 3, and especially level 2, of the strata profile which correspond to later episodesin the history of the quarry pit.

Change through time at the workshop

The environmental context of the workshop area on the south side of Maymeja has several characteristics commonly associated with residential base camp sites and these environmental features appear to have contributed to the concentration of obsidian production activities that occurred in this area. These site characteristics include perennial water nearby, grazing, long hours of direct sunlight with a relative abundance of shelter providing partial protection from wind.

The artifactual evidence from the A03-126 workshop, bracketed by three 14C dates between the Terminal Archaic and the Middle Formative, suggest that there were three principal episodes of activity at the workshop. First, in the early history of the workshop obsidian knappers took advantage of relatively large obsidian nodules that were apparently available in Maymeja. Next, the production pattern shifted and long, narrow obsidian cores of Ob1 material became abundant. Subsequently, after a brief ambiguous phase that perhaps reflected the exploitation of residual materials from the preceding stage, there was return to a pattern that is similar to the earliest occupation levels in terms of flake morphology and technical class variability. These episodes can be interpreted as demonstrating (1) initial local, irregular production, (2) intensive procurement and production that perhaps involved larger scale quarrying, (3) a return to irregular production and occasional quarrying. This evidence will be explored below.

Figure 7-11. Location of test unit Q02-2u3 in site A03-126 workshop.

Q02-2u3 Descriptive Characteristics by Level

The stratigraphy at the Maymeja workshop contains layers at two distinct angles that suggest that mounding occurred initially downslope of where the mound is located today. The Q02-2u3 test unit measured only 1x1m and provided little spatial evidence for architectural interpretation, but the yellow surface labeled as S3 is suggestive of some kind of prepared surface associated with obsidian production at the workshop. Radiocarbon samples recovered from levels 4, 6, and 7 indicate that the sequence in this test unit spans the middle of the Terminal Archaic through to the Early/Middle Formative and probably well into the Middle Formative. A soil sample examined by UC Santa Barbara volcanologists from level 6 contained grains of biotite and hornblende but these grains may have derived from the older bedrock (Sarah Fowler, pers. comm., 27 Feb 2006). The sampling strategy during excavation at the obsidian source involved collecting 25% of flakes being excavated per level by retaining flakes from a single quad, while carbon samples, cores and retouched artifacts (most of which were presumably observed by the excavator) were collected from throughout the 1x1m unit. Thus, flake counts and weights are proportionally ¼ of the amount of other artifact types collected from the Q02-2u3 excavation.

Figure 7-12. Q02-2u3 workshop test unit west and north profile diagrams showing strata.

Figure 7-13. Q02-2u3 west profile at Maymeja workshop site [A03-126].

Table 7-13. Q02-2u3 Excavation levels from test unit at the A03-126 workshop.

As with other units in Maymeja, TU3 consisted of almost exclusively obsidian artifacts in a matrix of tephra-rich soils. As would be expected at a workshop site, there was a great deal of redundancy both in material types and technical classes. Hammerstones were not found in abundance in the excavation, but cores made a large contribution to the assemblages in most levels. This is considerably more than are encountered in consumption sites. For example, in all 13 levels in the test unit A02-26u1 (Taukamayo) only 15 cores were found.

Figure 7-14. Proportions of Tech Classes in excavation levels by count in Q02-2u3.

Table 7-14. Q02-2u3:c²showing Technical Class by Level with the Standard Residual.

The variety of technical classes by level is shown above, with very significant counts of retouched flakes in level 6, and significantly low counts of retouched flakes in level 3. There are also a significantly high number of cores discarded in level 4. The weight of these assemblages is also important because, for example, while the count is high in level 3 the weight is relatively low as the assemblage consists of many small flakes. The thickness of each layer (HBD in Table 7-13) can be compared with weight.

Further examination of Q02-2u3 "workshop" sequence

Fluctuations in the presence of particular stone artifact types in the excavation test unit at the Maymeja workshop provides insights into changing reduction activities at the workshop area through time. As described above, radiocarbon dates for the unit place the lower parts of the excavation in the Terminal Archaic, while the middle of the sequence (around level 4) falls at the Early / Middle Formative transition. As was reviewed in Chapter 3, these are times of interest in terms of obsidian distributions throughout the region.

Connection to the quarry pit

An important question remains concerning the history of the Q02-2 quarry pit that lies 600m uphill to the east of the workshop. Due to a lack of datable materials or diagnostic artifacts in the quarry pit test unit Q02-2u2, it was not possible to directly link excavation levels at that unit with activities and excavation levels at the workshop test unit Q02-2u3 on a temporal basis.

If large nodules were being transported from the quarry to the site of A03-126 where they were being reduced, the morphology and visual characteristics of cortical flakes in Q02-2u3 may reflect this import of cores from the quarry pit up the hill. These tables can be compared with data presented earlier in Table 7-9 and Table 7-11 from the quarry pit test unit.

Table 7-25. Q02-2u3: Obsidian type and color of artifacts by level.

Table 7-26. Q02-2u3: Length of complete flakes and cores with ?50% cortex.

In comparing these data with the tables discussed earlier from the Q02-2u2 quarry test unit, no specific pattern is apparent that decisively links the quarry pit with the workshop downslope. An increase in Clear and Clear-Banded obsidian becomes visible in level 3 of the workshop (q02-2u3), in Table 7-10 shown above, that could reflect a sudden availability of this material from excavation at the quarry pit provided that the quarry pit is the sole source of this clear and clear banded obsidian. However as mentioned, the quarry pit is probably not the sole source of clear obsidian because it contains proportions of clear obsidian that reflect tool use over the entire region. There was concurrently a decrease in other colors of obsidian in level 3. No discrete pattern was observed in the Ob1 vs Ob2 material type linking the two pits. Finally, the length of cortical flakes indicates that dramatically more long flakes were deposited at the source from level 5 onwards, as discussed earlier, but there were also notably more medium and small cortical flakes in those levels as well.

Discussion

Evidence from TU2 (A02-39u2] suggests that this secondary rock oval structure [A03-557] to the south-west of the main Pausa ovals, was perhaps an overflow area. As the test unit was placed at the bottom-most portion of the oval (presumably a corral structure) much of the bone and charcoal that was encountered was perhaps fill that was swept to the lowest point in the corral by animal activity. It appears that during the Early Tiwanaku period the enclosed area was relatively level and charcoal and bone were swept to the bottom of the corral area.

Stratigraphy

The local test unit datum for both of these units was placed on a large, flat rock that is part of the A03-558 oval that lies just to the south of TU3. This mapping station, as well as the site datum north, are shown in Figure 7-32 in the positions where they were mapped with a total station in 2004.

Table 7-27. A02-39u3 test unit. Excavation levels, overview, and soil description.

Table 7-28. A02-39u4 test unit. Excavation levels, overview, and soil description.

Figure 7-33. A02-39u4 Pausa test unit west profile diagram showing strata and levels.

Radiocarbon dates recovered from levels 3 and 5 of Unit 4 are temporally inverted, but they overlap in the 2-sigma error ranges. These dates place the later use of the hearth in the vicinity of AD200, or the Upper Formative. At units 3 and 4 at Pausa showed a relatively high percentage of obsidian by count throughout the sequence (83%), although by weight obsidian was somewhat less dominant (58%), while volcanics became more prominent (33%).

Artifacts by level

Table 7-29. A03-39u3/u4 counts of lithics by level.

Figure 7-34. A02-39u3/u4 Bar graph showing counts of lithics classes by excavation level.

Classes of lithics from A02-39u3 and u4 show that demonstrable variability was occurring in activities in this area. All of the diagnostic projectile points were Series 5 points, all points were made from Ob1 obsidian except for one Ob2 point, and 80% of the points were broken. The complete points were found scattered in every level, and the series 5 types are consistent with the Late Formative dates from radiocarbon dating. Some of the points were extremely small, including one beautifully pressure flaked point [L03-60.3] that measured only 12.9mm long by 5.2mm wide, and 2.9mm thick weighing 0.1 g. The other points were all very small, with a mean weight of 0.6 grams, which suggests that the size of the projectile is not necessarily related to material abundance as obsidian was abundant at this location in the Late Formative and yet the projectile points were extremely small.

Table 7-30. A02-39u3 and u4: Obsidian flake types by excavation level.

In level 5, between 50 and 60 cm depth and associated with the F1 hearth feature in south half of Unit 4, a distinct increase in obsidian reduction activity was encountered that suggests more advanced reduction activity was occurring in this period. Focusing only on complete obsidian flakes, reveals a much higher count of complete flakes per unit volume of dirt, a quantity that was more than double the number in other levels.

Leaving out level 1 due to surface contamination, the evidence from complete flakes suggests that while bifacial reduction was occurring in level 5, pressure flaking was relatively rare. Bifacial thinning flakes were noted by one of the analysts (Mackay) in the process of labwork (observed BTF column) but because other analysts were not making note of BTF during analysis, additional candidates for inclusion in the BTF category based on the BTF calculation model presented above appear in the column labeled "Calculated BTF". Both observed BTF and calculated BTF values suggest that during level 5 distinctive reduction activities occurred.

Figure 7-35. Cumulative Frequency of A02-39 u3 & u4 showing Level 5 reduction against reduction for all levels.

Figure 7-35 shows that there was a larger number of very small flakes in level 5, where over 75% of the complete flakes weighed less than 0.6 grams. Evidence of the conservation of material appears in levels 4 and 5, where complete flakes have a higher mean number of rotations and much smaller platform sizes. Finally, the mean coverage of dorsal cortex on flakes decreased from 6.9% in level 5 to 4.3% in level 4, while 6.4% is the mean for these units. In other words, there was a slightly greater range of reduction levels occurring in level 5 where there was more cortex but also a great many small flakes. By the next level, level 4, there was primarily decorticated material in use, and relatively little cortical obsidian.

Other artifacts include bone and ceramics throughout the sequence, and a few tiny pieces of turquoise in level 6 at the bottom of the hearth. Ceramics were found in every level in these excavation units, but few of them were diagnostic. Of the 38 sherds only four were rim sherds and two were handle pieces. One sherd found in level 5 had an incised line parallel to the rim, but was otherwise non-diagnostic. Five sherds from the lower levels (level 3, 5, and 7) resemble the "Chiquero" style described by Wernke (2003) with a brushed interior and no slip. Bone fragments were found in every level but none were diagnostic. Burned bone and charcoal in grey soil were found in particularly high density in hearth F1, along with a broken piece of groundstone and fire-cracked rock.

Discussion

Evidence from these test units suggest that the small hearth was built between windscreens and walls made of standing slab rocks. Based on the small size of the hearth and the structure size, the hearth seems to have been a residential feature. Assuming the overall site configuration at Pausa involved two large ovals of rocks that may be observed today, the structure that was encountered in unit 3 and 4 pits would be immediately north of the internal oval. These ovals may have been herd control features, such as large corrals with well constructed wall bases on slightly raised mounds, and this residential structure would have been constructed against the north perimeter of the inner circular corral feature. Evidence from lithic reduction show that advanced stages of lithic reduction were occurring here, and while the evidence of bifacial production and pressure flaking absent from the Chivay workshop area, such activities appear to have been occurring with greater frequency in Block 2.

Stratigraphy

Inspection of the slope cut and the stratigraphy of the test unit confirmed that the site had not slid in the past and the stratigraphy was largely intact. There are signs of increasing angle of deposition through time. Sloping, colluviated soils gave way to level, well-sorted gravels near the bottom of the cultural levels. Radiocarbon dating indicated that the deposit is largely synchronic as the two sigma error of the radiocarbon samples overlap.

Figure 7-36. Test unit A02-26u1 is a 1x1m test unit placed on the edge of a creeping landslide with an irregular extension area (1x) on the south edge of the unit. Top of the photo is north.

The TU1 was placed immediately north of the cut slope, but because it was oriented to true north there was a small irregular, extra section to the south of the unit that dubbed A02-26u1 x. As a cut profile was available it was possible to observe that the ~60cm of overburden above the charcoal, bone, lithic, and ceramic layer in the profile was virtually sterile. The overburden was excavated relatively expediently by only screening only one-quarter of the dirt through 1/16" screen, the rest went through 1/4" screen only. By level 3 (58cm below datum) everything was screened through 16" screen.

Table 7-31. A02-26u1 test unit. Excavation levels, overview, and soil description.

Figure 7-37. A02-26u1 Taukamayo test unit, north profile diagram showing strata and levels.

The stratigraphy at A02-26u1 revealed that the site was intact behind the profile as no obvious landslide had disturbed the soils. There was a distinct change in angle approximately 60-80cm below the surface where the strata are at a much lower angle. These strata contain a great deal of charcoal and bone, and the radiocarbon dates place the assemblage at approximately AD700. This is the middle of the Tiwanaku period and it represents the beginning of the Middle Horizon. The slope angle change that is suggested by the soil stratigraphy is likely related to changes in the position or altitude of the adjacent creek. Evidence of alluvial sorting and river smoothed stones were also found in strata below the S4 and S5 occupation, which suggest that these occupations are on top of an old river terrace. Today, Taukamayo creek is approximately 40m vertical below the site altitude, and denuded hillslopes around the creek suggest that erosion is severe in this tributary valley to the Colca. A preliminary assessment is that Taukamayo creek has significantly incised since the Late Formative and Middle Horizon when it flowed below the site of A02-26 which lay on a relatively flat T1 terrace. Increased incising of Taukamayo creek first caused soil deposition to assume steeper angles, as is apparent in the upper 60cm of the profile. However in the more recent past the creek has begun to undercut the T1 terrace on which the site rests and the southern part of Taukamayo has since slid, revealing the profile cut.

Artifacts by level

Obsidian arrives into the test unit area in conjunction with bones, both burned and unburned, that turned out to be primarily camelid bones. The lithic analysis evidence suggests that obsidian was brought into the area around level 8 where the mean flake size is considerably higher than average and obsidian goes from an average 55% of the lithic assemblage by level, to 80% of the assemblage in level 8 and then 70% of the assemblage in level 7.

Table 7-32. A02-26u1: Artifact counts and averages for select measures.

Table 7-33. A02-26u1: Lithic Technical Classes by Level showing counts.

Figure 7-38. A02-26u1 Technical Class by Excavation Level.

Figure 7-39. A02-26u1 Material Types by Excavation Level.

Figure 7-40. Obsidian at Taukamayo A02-26u1, Ob1 and Ob2 compared.

Material types at Taukamayo provide insights regarding the use of non-obsidian lithic material in the vicinity of the Chivay source. Chert is abundant in the Colca and the Llapa rivers, and in Taukamayo creek. Yet the reduction in imported obsidian appears to have occurred at a few distinct sites in the Upper Colca valley area, as if complementing the local chert use.

The use of Ob1 and Ob2 obsidian at Taukamayo is also informative. The influx of obsidian in levels 6-8 appears to have been from a source with Ob2 obsidian as both types of obsidian increase in these levels but the percentage of Ob2 obsidian is significantly greater than it was in previous and later levels. In the course of survey work, Ob2 obsidian was encountered outside of the Maymeja zone in Block 4, and also at a minor obsidian source in the Pulpera drainage called Condorquiña (Section 4.5.1). It is possible that the Ob2 obsidian was derived from one of these sources, and that this lower quality obsidian was generally not intended for projectile point production. In fact, projectile points are comparatively rare at Taukamayo, with only 8 obsidian projectile points and preforms coming from the TU1 pit, and of these only one point (in level 7) was Ob2 material. This evidence suggests that the focus on obsidian production in this area in the beginning of the Middle Horizon was on flake and biface production, and there was relatively little preference shown for Ob1 obsidian over Ob2 material.

In level 5, there appears to have been more advanced reduction of obsidian occurring. The BTF index suggests that bifacial thinning flakes are more prevalent than in other levels, and 3 broken bifaces (50% of total) were found in level 5, the average number of rotations went from 0.36 to 0.56 in level 5 implying that greater conservation of material was occurring in that level.

Figure 7-41. Cumulative frequency comparing obsidian and non-obsidian flakes.

Levels 5 and 6 also contained a high density of animal bone. Analysis of diagnostic faunal remains by Randi Gladwell revealed that the majority of the bones were camelid, and though some small mammal bones were encountered that may have been cuy(guinea pig) or viscachabones. Comparing obsidian with non-obsidian overall, it is apparent that more advanced reduction occurred on obsidian. This contrast is expressed in Figure 7-41 where over 80% of obsidian flakes are under 1.7g in weight.

Comparing Material Types in A02-26u1 and Block 3 overall

The excavation unit A02-26u1 at Taukamayo presents lithic evidence, albeit synchronic to the Tiwanaku period, that is, in some ways, representative of the material use pattern in Block 3 overall. Distinct patterns emerge, however, with respect to the use of Ob2 obsidian, because as was noted Ob2 obsidian was rarely used outside of Block 3.

Table 7-34. A02-26u1 Excavation and Block 3 surface: Lithic Tech. Classes by Obsidian type.

Table 7-34 shows that Ob1 obsidian was primarily used for projectile point production, however bifaces from Ob2 were used to a much greater extent at A02-26u1. Retouched flakes of the two material types were used in similar proportions in the two contexts.

Discussion

This test unit appears to show several occasions of obsidian reduction in conjunction with camelid butchering and cooking during the earliest part of the Tiwanaku period at this site in the upper Colca. More advanced reduction was occurring on obsidian, but there appeared to have been little preference for Ob1 obsidian over the Ob2 material, and in fact in the production of bifaces it appears knappers used Ob2 preferentially, though both materials were available. The Uyo Uyo obsidian source is under 40 km downstream and there are irregularities in Uyo Uyo material that such that it falls in the Ob2 category.

As dates contemporary with Tiwanaku were derived from this unit, a regional consideration of this occupation should include the distributions of Chivay obsidian in this time period. Regional distributions indicate that during the Tiwanaku period Chivay obsidian was rarely used north of Chivay source area. Distributions tend strongly towards the south and east where the Lake Titicaca Basin consumption pattern is strongest. Yet it is curious that the highest density of obsidian in the Block 3 area north of the source is dated to the Tiwanaku period, which is in a northward direction and towards the Wari sphere. In concordance with regional obsidian distributions in this time period, it is notable that the heavy use of Ob2 obsidian, and the production of bifaces rather than projectile points, suggests that goods produced at this obsidian production area use were intended for modest local consumption.